fix [redundant_closure] suggesting incorrect code with F: Fn()

[J-ZhengLi]

fixes: #12853

---

changelog: fix [redundant_closure] suggesting incorrect code with F: Fn()

[rustbot]

r? @y21

rustbot has assigned @y21.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[J-ZhengLi]

confused, but too scared to remove it

[y21]

Yeah, I'm also not entirely sure what that is for. Seems like that additional check was introduced in #8685, but that was a rather large refactor. Before this, there was only the local_used_after_expr check. Unfortunate that there's no test for this.

[J-ZhengLi]

Oh, it might mean situation like this?

let fun = |x: i32| -> i32 {
    x
};

let _ = fun(fun(0));

but I'm not sure, I'll test it.

[y21]

I don't think so. There is a check somewhere in the if chain that makes sure that the fn arguments are only paths to locals. So in your example that wouldn't happen because fun(0) is more than just a path.

rust-clippy/clippy_lints/src/eta_reduction.rs

Lines 128 to 131 in 03654ba

	|| !check_inputs(typeck, body.params, None, args)
	{
	return;
	}

rust-clippy/clippy_lints/src/eta_reduction.rs

Lines 230 to 234 in 03654ba

	matches!(
	p.pat.kind,
	PatKind::Binding(BindingMode::NONE, id, _, None)
	if path_to_local_id(arg, id)
	)

[J-ZhengLi]

right... then I'm afraid that I can't think of any examples that pass the compilation check😂

[y21]

Don't think we need to block it on figuring out what that local_used_in check is for. LGTM, thanks!

@bors r+

[bors]

📌 Commit db30f6c has been approved by y21

It is now in the queue for this repository.

[bors]

⌛ Testing commit db30f6c with merge 0b598b6...

[bors]

☀️ Test successful - checks-action_dev_test, checks-action_remark_test, checks-action_test
Approved by: y21
Pushing 0b598b6 to master...